Analyses for ‘Seasonal variation of micro and macro soil invertebrate communities in a temperate forest’
2024-March-10
Across Year Effects
How does the seasonal variability impact different invertebrate communities?
Climate Variables
Nematodes:
Total Nematodes
Significant drivers were season and the interaction between topography and tree functional type.
model.total <-lme(total.nematodes~season+landscape.position+mycorrhizal.fungi.type+season:landscape.position+
mycorrhizal.fungi.type:landscape.position+season:mycorrhizal.fungi.type+
season:landscape.position:mycorrhizal.fungi.type,
random=list(sample.id=~1, tree.identity=~1), data = nematode)
summary(model.total)$tTable Value Std.Error DF t-value p-value
(Intercept) 4752.9 3136 100 1.5156 0.1328
seasonSpring 403.3 3489 100 0.1156 0.9082
seasonSummer 5744.1 3748 100 1.5327 0.1285
seasonFall 8627.2 5464 29 1.5790 0.1252
seasonWinter.early 13463.7 4455 29 3.0224 0.0052
landscape.positionUphill 10449.8 4945 29 2.1134 0.0433
mycorrhizal.fungi.typeECM 5091.7 6213 29 0.8195 0.4192
seasonSpring:landscape.positionUphill -5268.3 5450 100 -0.9667 0.3360
seasonSummer:landscape.positionUphill -2879.1 5833 100 -0.4936 0.6227
seasonFall:landscape.positionUphill 683.2 7602 29 0.0899 0.9290
seasonWinter.early:landscape.positionUphill -85.9 6668 29 -0.0129 0.9898
landscape.positionUphill:mycorrhizal.fungi.typeECM -9308.2 8237 29 -1.1301 0.2677
seasonSpring:mycorrhizal.fungi.typeECM -2131.9 6696 100 -0.3184 0.7508
seasonSummer:mycorrhizal.fungi.typeECM -3373.5 7089 100 -0.4759 0.6352
seasonFall:mycorrhizal.fungi.typeECM -4380.6 8102 29 -0.5407 0.5929
seasonWinter.early:mycorrhizal.fungi.typeECM -10537.9 7656 29 -1.3764 0.1792
seasonSpring:landscape.positionUphill:mycorrhizal.fungi.typeECM 4143.7 8980 100 0.4614 0.6455
seasonSummer:landscape.positionUphill:mycorrhizal.fungi.typeECM 6900.9 9506 100 0.7260 0.4696
seasonFall:landscape.positionUphill:mycorrhizal.fungi.typeECM -2226.4 10911 29 -0.2041 0.8397
seasonWinter.early:landscape.positionUphill:mycorrhizal.fungi.typeECM 1985.3 10386 29 0.1911 0.8497anova(model.total) numDF denDF F-value p-value
(Intercept) 1 92 371 <.0001
season 4 92 9 <.0001
landscape.position 1 37 15 0.0004
mycorrhizal.fungi.type 1 37 3 0.0811
season:landscape.position 4 92 1 0.7350
landscape.position:mycorrhizal.fungi.type 1 37 4 0.0468
season:mycorrhizal.fungi.type 4 92 2 0.1226
season:landscape.position:mycorrhizal.fungi.type 4 92 0 0.8527AIC(model.total)[1] 2856rsquared(model.total) Response family link method Marginal Conditional
1 total.nematodes gaussian identity none 0.381 0.492pairs(emmeans(model.total, specs= ~landscape.position*mycorrhizal.fungi.type)) contrast estimate SE df t.ratio p.value
Downhill AM - Uphill AM -8940 2140 29 -4.190 0.0010
Downhill AM - Downhill ECM -1007 2280 29 -0.440 0.9710
Downhill AM - Uphill ECM -2799 2100 29 -1.340 0.5490
Uphill AM - Downhill ECM 7933 2310 29 3.430 0.0090
Uphill AM - Uphill ECM 6141 2120 29 2.890 0.0340
Downhill ECM - Uphill ECM -1792 2270 29 -0.790 0.8590
Results are averaged over the levels of: season
Degrees-of-freedom method: containment
P value adjustment: tukey method for comparing a family of 4 estimates 
Total Nematode Abundance Time Series
Bacterial Feeding Nematodes
Bacterial feeder abundance was best explained by season.
model.bf <-lme(bacterial.feeders~season+landscape.position+season:landscape.position+
season:mycorrhizal.fungi.type+season:landscape.position:mycorrhizal.fungi.type,
random=list(sample.id=~1, tree.identity=~1), data = nematode)
summary(model.bf)$tTable Value Std.Error DF t-value p-value
(Intercept) 2128.2 1590 98 1.3383 0.184
seasonSpring 668.1 1754 98 0.3809 0.704
seasonSummer 2490.2 1901 98 1.3101 0.193
seasonFall 2289.9 2788 31 0.8214 0.418
seasonWinter.early 3546.2 2269 31 1.5626 0.128
landscape.positionUphill 2105.8 2506 31 0.8402 0.407
seasonSpring:landscape.positionUphill -2114.0 2741 98 -0.7712 0.442
seasonSummer:landscape.positionUphill -1093.8 2957 98 -0.3700 0.712
seasonFall:landscape.positionUphill 758.8 3871 31 0.1960 0.846
seasonWinter.early:landscape.positionUphill -3053.8 3395 31 -0.8996 0.375
seasonWinter.late:mycorrhizal.fungi.typeECM 1979.2 3147 98 0.6289 0.531
seasonSpring:mycorrhizal.fungi.typeECM 748.1 2025 98 0.3694 0.713
seasonSummer:mycorrhizal.fungi.typeECM 63.3 1844 98 0.0343 0.973
seasonFall:mycorrhizal.fungi.typeECM -404.5 2661 31 -0.1520 0.880
seasonWinter.early:mycorrhizal.fungi.typeECM -2564.0 2290 31 -1.1199 0.271
seasonWinter.late:landscape.positionUphill:mycorrhizal.fungi.typeECM -2340.7 4173 98 -0.5609 0.576
seasonSpring:landscape.positionUphill:mycorrhizal.fungi.typeECM -1232.8 2802 98 -0.4400 0.661
seasonSummer:landscape.positionUphill:mycorrhizal.fungi.typeECM 2209.2 2547 98 0.8673 0.388
seasonFall:landscape.positionUphill:mycorrhizal.fungi.typeECM -2955.5 3656 31 -0.8084 0.425
seasonWinter.early:landscape.positionUphill:mycorrhizal.fungi.typeECM 2146.4 3238 31 0.6629 0.512anova(model.bf) numDF denDF F-value p-value
(Intercept) 1 90 147.6 <.0001
season 4 90 3.0 0.0232
landscape.position 1 39 1.7 0.2016
season:landscape.position 4 90 0.7 0.5631
season:mycorrhizal.fungi.type 5 90 0.6 0.6969
season:landscape.position:mycorrhizal.fungi.type 5 90 0.5 0.7868AIC(model.bf)[1] 2670rsquared(model.bf) Response family link method Marginal Conditional
1 bacterial.feeders gaussian identity none 0.148 0.312pairs(emmeans(model.bf, specs= ~season))NOTE: A nesting structure was detected in the fitted model:
mycorrhizal.fungi.type %in% seasonNOTE: Results may be misleading due to involvement in interactions contrast estimate SE df t.ratio p.value
Winter.late - Spring 727 1130 98 0.640 0.9670
Winter.late - Summer -2123 1200 98 -1.760 0.4010
Winter.late - Fall -1324 1390 31 -0.950 0.8730
Winter.late - Winter.early -869 1320 31 -0.660 0.9640
Spring - Summer -2850 866 98 -3.290 0.0120
Spring - Fall -2051 1150 31 -1.780 0.4020
Spring - Winter.early -1597 1070 31 -1.490 0.5750
Summer - Fall 799 1110 31 0.720 0.9510
Summer - Winter.early 1253 1030 31 1.220 0.7420
Fall - Winter.early 454 1220 31 0.370 0.9960
Results are averaged over the levels of: landscape.position, mycorrhizal.fungi.type
Degrees-of-freedom method: containment
P value adjustment: tukey method for comparing a family of 5 estimates Fungal Feeding Nematodes
Fungal feeder abundance was best explained by season.
model.ff <-lme(fungal.feeders~season+landscape.position+ season:landscape.position+mycorrhizal.fungi.type:landscape.position+
+season:landscape.position:mycorrhizal.fungi.type,
random=list(sample.id=~1, tree.identity=~1), data = nematode)
summary(model.ff)$tTable Value Std.Error DF t-value p-value
(Intercept) 74.243 176 100 0.42086 0.675
seasonSpring 0.156 176 100 0.00089 0.999
seasonSummer 47.931 209 100 0.22927 0.819
seasonFall 208.031 332 29 0.62594 0.536
seasonWinter.early 112.956 266 29 0.42455 0.674
landscape.positionUphill 79.329 277 29 0.28675 0.776
seasonSpring:landscape.positionUphill -195.584 276 100 -0.70935 0.480
seasonSummer:landscape.positionUphill -43.881 323 100 -0.13596 0.892
seasonFall:landscape.positionUphill 570.657 451 29 1.26398 0.216
seasonWinter.early:landscape.positionUphill -199.457 395 29 -0.50520 0.617
landscape.positionDownhill:mycorrhizal.fungi.typeECM 340.494 343 29 0.99171 0.330
landscape.positionUphill:mycorrhizal.fungi.typeECM -146.584 301 29 -0.48642 0.630
seasonSpring:landscape.positionDownhill:mycorrhizal.fungi.typeECM -224.593 342 100 -0.65584 0.513
seasonSummer:landscape.positionDownhill:mycorrhizal.fungi.typeECM -217.399 386 100 -0.56270 0.575
seasonFall:landscape.positionDownhill:mycorrhizal.fungi.typeECM -47.806 474 29 -0.10088 0.920
seasonWinter.early:landscape.positionDownhill:mycorrhizal.fungi.typeECM -245.604 444 29 -0.55304 0.584
seasonSpring:landscape.positionUphill:mycorrhizal.fungi.typeECM 197.016 306 100 0.64323 0.522
seasonSummer:landscape.positionUphill:mycorrhizal.fungi.typeECM 6.206 350 100 0.01776 0.986
seasonFall:landscape.positionUphill:mycorrhizal.fungi.typeECM -384.889 425 29 -0.90594 0.372
seasonWinter.early:landscape.positionUphill:mycorrhizal.fungi.typeECM 324.051 412 29 0.78559 0.438anova(model.ff) numDF denDF F-value p-value
(Intercept) 1 92 21.30 <.0001
season 4 92 4.28 0.0032
landscape.position 1 37 0.28 0.6008
season:landscape.position 4 92 0.31 0.8731
landscape.position:mycorrhizal.fungi.type 2 37 1.08 0.3497
season:landscape.position:mycorrhizal.fungi.type 8 92 0.56 0.8041AIC(model.ff) [1] 2056rsquared(model.ff) Response family link method Marginal Conditional
1 fungal.feeders gaussian identity none 0.213 0.479pairs(emmeans(model.ff, specs= ~season))NOTE: A nesting structure was detected in the fitted model:
mycorrhizal.fungi.type %in% landscape.positionNOTE: Results may be misleading due to involvement in interactions contrast estimate SE df t.ratio p.value
Winter.late - Spring 105 115.0 100 0.910 0.8920
Winter.late - Summer 27 130.0 100 0.210 1.0000
Winter.late - Fall -385 159.0 29 -2.420 0.1380
Winter.late - Winter.early -33 152.0 29 -0.220 0.9990
Spring - Summer -78 91.5 100 -0.850 0.9140
Spring - Fall -490 136.0 29 -3.610 0.0090
Spring - Winter.early -137 127.0 29 -1.080 0.8140
Summer - Fall -412 134.0 29 -3.080 0.0330
Summer - Winter.early -60 124.0 29 -0.480 0.9890
Fall - Winter.early 352 149.0 29 2.370 0.1540
Results are averaged over the levels of: mycorrhizal.fungi.type, landscape.position
Degrees-of-freedom method: containment
P value adjustment: tukey method for comparing a family of 5 estimates Plant Parasitic Nematodes
Best explained by season and the interaction between topography and tree functional type.
model.pp <-lme(plant.parasites.aph.~season+landscape.position+mycorrhizal.fungi.type + season:landscape.position+mycorrhizal.fungi.type:landscape.position+
season:mycorrhizal.fungi.type+season:landscape.position:mycorrhizal.fungi.type,
random=list(sample.id=~1, tree.identity=~1), data = nematode)
summary(model.pp)$tTable Value Std.Error DF t-value p-value
(Intercept) 2143 2322 100 0.9228 0.3583
seasonSpring -174 2614 100 -0.0665 0.9471
seasonSummer 2831 2774 100 1.0207 0.3098
seasonFall 5535 4010 29 1.3802 0.1781
seasonWinter.early 8656 3277 29 2.6417 0.0132
landscape.positionUphill 7295 3663 29 1.9918 0.0559
mycorrhizal.fungi.typeECM 2318 4607 29 0.5031 0.6187
seasonSpring:landscape.positionUphill -2363 4082 100 -0.5790 0.5639
seasonSummer:landscape.positionUphill -1480 4320 100 -0.3426 0.7326
seasonFall:landscape.positionUphill -606 5597 29 -0.1084 0.9145
seasonWinter.early:landscape.positionUphill 3052 4910 29 0.6215 0.5391
landscape.positionUphill:mycorrhizal.fungi.typeECM -5553 6106 29 -0.9094 0.3706
seasonSpring:mycorrhizal.fungi.typeECM -358 5008 100 -0.0714 0.9432
seasonSummer:mycorrhizal.fungi.typeECM -632 5259 100 -0.1203 0.9045
seasonFall:mycorrhizal.fungi.typeECM -1480 5973 29 -0.2478 0.8060
seasonWinter.early:mycorrhizal.fungi.typeECM -5279 5650 29 -0.9343 0.3578
seasonSpring:landscape.positionUphill:mycorrhizal.fungi.typeECM 1735 6716 100 0.2584 0.7967
seasonSummer:landscape.positionUphill:mycorrhizal.fungi.typeECM 1166 7048 100 0.1655 0.8689
seasonFall:landscape.positionUphill:mycorrhizal.fungi.typeECM -2129 8047 29 -0.2645 0.7932
seasonWinter.early:landscape.positionUphill:mycorrhizal.fungi.typeECM -3191 7659 29 -0.4166 0.6800anova(model.pp) numDF denDF F-value p-value
(Intercept) 1 92 231.6 <.0001
season 4 92 7.5 <.0001
landscape.position 1 37 14.8 0.0005
mycorrhizal.fungi.type 1 37 4.7 0.0370
season:landscape.position 4 92 0.5 0.7535
landscape.position:mycorrhizal.fungi.type 1 37 7.1 0.0115
season:mycorrhizal.fungi.type 4 92 1.8 0.1413
season:landscape.position:mycorrhizal.fungi.type 4 92 0.2 0.9276pairs(emmeans(model.pp, specs= ~landscape.position*mycorrhizal.fungi.type)) contrast estimate SE df t.ratio p.value
Downhill AM - Uphill AM -7016 1560 29 -4.490 0.0010
Downhill AM - Downhill ECM -768 1670 29 -0.460 0.9670
Downhill AM - Uphill ECM -1747 1530 29 -1.140 0.6690
Uphill AM - Downhill ECM 6247 1690 29 3.690 0.0050
Uphill AM - Uphill ECM 5268 1560 29 3.390 0.0100
Downhill ECM - Uphill ECM -979 1670 29 -0.590 0.9350
Results are averaged over the levels of: season
Degrees-of-freedom method: containment
P value adjustment: tukey method for comparing a family of 4 estimates AIC(model.pp)[1] 2777rsquared(model.pp) Response family link method Marginal Conditional
1 plant.parasites.aph. gaussian identity none 0.38 0.481pairs(emmeans(model.pp, specs= ~season)) contrast estimate SE df t.ratio p.value
Winter.late - Spring 1100 1680 100 0.660 0.9650
Winter.late - Summer -2066 1760 100 -1.170 0.7670
Winter.late - Fall -3959 2010 29 -1.970 0.3060
Winter.late - Winter.early -6745 1920 29 -3.520 0.0120
Spring - Summer -3167 1280 100 -2.480 0.1030
Spring - Fall -5060 1660 29 -3.050 0.0370
Spring - Winter.early -7845 1540 29 -5.090 <.0001
Summer - Fall -1893 1600 29 -1.180 0.7610
Summer - Winter.early -4679 1480 29 -3.170 0.0270
Fall - Winter.early -2786 1750 29 -1.590 0.5130
Results are averaged over the levels of: landscape.position, mycorrhizal.fungi.type
Degrees-of-freedom method: containment
P value adjustment: tukey method for comparing a family of 5 estimates Omnivorous and Predatory Nematodes
A new group created with the sum of omnivorous and predatory nematodes. This group was best explained by season and topography.
model.pred1 <-lme(om.pr~ season+landscape.position+ mycorrhizal.fungi.type + season:landscape.position + mycorrhizal.fungi.type:landscape.position + season:mycorrhizal.fungi.type + season:mycorrhizal.fungi.type:landscape.position,
random=list(sample.id=~1, tree.identity=~1), data = nematode)
summary(model.pred1)$tTable Value Std.Error DF t-value p-value
(Intercept) 175.8 204 100 0.8633 0.3900
seasonSpring -29.7 257 100 -0.1156 0.9082
seasonSummer 359.4 239 100 1.5037 0.1358
seasonFall 417.8 323 29 1.2942 0.2058
seasonWinter.early 623.4 270 29 2.3097 0.0282
landscape.positionUphill 402.2 322 29 1.2492 0.2216
mycorrhizal.fungi.typeECM 186.6 407 29 0.4582 0.6502
seasonSpring:landscape.positionUphill -155.5 398 100 -0.3903 0.6971
seasonSummer:landscape.positionUphill 260.8 374 100 0.6965 0.4877
seasonFall:landscape.positionUphill -19.9 465 29 -0.0428 0.9662
seasonWinter.early:landscape.positionUphill 386.5 408 29 0.9474 0.3513
landscape.positionUphill:mycorrhizal.fungi.typeECM -450.5 539 29 -0.8362 0.4099
seasonSpring:mycorrhizal.fungi.typeECM -87.9 481 100 -0.1828 0.8553
seasonSummer:mycorrhizal.fungi.typeECM -277.2 461 100 -0.6010 0.5492
seasonFall:mycorrhizal.fungi.typeECM 81.7 501 29 0.1629 0.8717
seasonWinter.early:mycorrhizal.fungi.typeECM -230.9 478 29 -0.4829 0.6328
seasonSpring:landscape.positionUphill:mycorrhizal.fungi.typeECM 448.9 645 100 0.6955 0.4884
seasonSummer:landscape.positionUphill:mycorrhizal.fungi.typeECM 446.0 616 100 0.7246 0.4704
seasonFall:landscape.positionUphill:mycorrhizal.fungi.typeECM 48.2 679 29 0.0709 0.9440
seasonWinter.early:landscape.positionUphill:mycorrhizal.fungi.typeECM -72.0 645 29 -0.1117 0.9118anova(model.pred1) numDF denDF F-value p-value
(Intercept) 1 92 317 <.0001
season 4 92 12 <.0001
landscape.position 1 37 25 <.0001
mycorrhizal.fungi.type 1 37 0 0.488
season:landscape.position 4 92 2 0.141
landscape.position:mycorrhizal.fungi.type 1 37 2 0.189
season:mycorrhizal.fungi.type 4 92 1 0.538
season:landscape.position:mycorrhizal.fungi.type 4 92 1 0.736AIC(model.pred1) #2145[1] 2145AICc(model.pred1) #2153[1] 2153rsquared(model.pred1) #37 Response family link method Marginal Conditional
1 om.pr gaussian identity none 0.365 0.368pairs(emmeans(model.pred1, specs= ~season))NOTE: Results may be misleading due to involvement in interactions contrast estimate SE df t.ratio p.value
Winter.late - Spring 39 161 100 0.240 0.9990
Winter.late - Summer -463 154 100 -3.010 0.0270
Winter.late - Fall -461 170 29 -2.710 0.0760
Winter.late - Winter.early -683 161 29 -4.240 0.0020
Spring - Summer -502 116 100 -4.320 <.0001
Spring - Fall -500 137 29 -3.650 0.0080
Spring - Winter.early -722 126 29 -5.730 <.0001
Summer - Fall 2 127 29 0.010 1.0000
Summer - Winter.early -221 116 29 -1.910 0.3370
Fall - Winter.early -222 136 29 -1.630 0.4890
Results are averaged over the levels of: landscape.position, mycorrhizal.fungi.type
Degrees-of-freedom method: containment
P value adjustment: tukey method for comparing a family of 5 estimates 
Nematode Trophic Groups by Season
Nematode Diversity (Simpson Diversity Index)
Simpson diversity was best explained by tree functional type and season
worm.simp1 <-lme(simpson~season+landscape.position+mycorrhizal.fungi.type +
season:landscape.position+mycorrhizal.fungi.type:landscape.position+
season:mycorrhizal.fungi.type+season:landscape.position:mycorrhizal.fungi.type,
random=list(sample.id=~1, tree.identity=~1), data = nematode)
summary(worm.simp1)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.51920 0.0386 100 13.4546 3.60e-24
## seasonSpring -0.01228 0.0482 100 -0.2547 7.99e-01
## seasonSummer 0.00101 0.0454 100 0.0223 9.82e-01
## seasonFall 0.00670 0.0616 29 0.1088 9.14e-01
## seasonWinter.early -0.09316 0.0514 29 -1.8119 8.04e-02
## landscape.positionUphill -0.04870 0.0610 29 -0.7982 4.31e-01
## mycorrhizal.fungi.typeECM 0.05998 0.0772 29 0.7772 4.43e-01
## seasonSpring:landscape.positionUphill -0.01960 0.0748 100 -0.2622 7.94e-01
## seasonSummer:landscape.positionUphill 0.04781 0.0711 100 0.6724 5.03e-01
## seasonFall:landscape.positionUphill 0.01313 0.0885 29 0.1483 8.83e-01
## seasonWinter.early:landscape.positionUphill 0.03137 0.0777 29 0.4039 6.89e-01
## landscape.positionUphill:mycorrhizal.fungi.typeECM -0.05196 0.1021 29 -0.5089 6.15e-01
## seasonSpring:mycorrhizal.fungi.typeECM -0.01799 0.0904 100 -0.1989 8.43e-01
## seasonSummer:mycorrhizal.fungi.typeECM -0.05997 0.0875 100 -0.6852 4.95e-01
## seasonFall:mycorrhizal.fungi.typeECM -0.05106 0.0954 29 -0.5353 5.97e-01
## seasonWinter.early:mycorrhizal.fungi.typeECM -0.01328 0.0907 29 -0.1465 8.85e-01
## seasonSpring:landscape.positionUphill:mycorrhizal.fungi.typeECM 0.06299 0.1214 100 0.5190 6.05e-01
## seasonSummer:landscape.positionUphill:mycorrhizal.fungi.typeECM 0.07877 0.1168 100 0.6742 5.02e-01
## seasonFall:landscape.positionUphill:mycorrhizal.fungi.typeECM 0.08175 0.1292 29 0.6329 5.32e-01
## seasonWinter.early:landscape.positionUphill:mycorrhizal.fungi.typeECM 0.09090 0.1225 29 0.7421 4.64e-01anova(worm.simp1)## numDF denDF F-value p-value
## (Intercept) 1 92 4242 <.0001
## season 4 92 3 0.0106
## landscape.position 1 37 1 0.2702
## mycorrhizal.fungi.type 1 37 5 0.0341
## season:landscape.position 4 92 1 0.4118
## landscape.position:mycorrhizal.fungi.type 1 37 0 0.5689
## season:mycorrhizal.fungi.type 4 92 0 0.8001
## season:landscape.position:mycorrhizal.fungi.type 4 92 0 0.9605AIC(worm.simp1)## [1] -172rsquared(worm.simp1)## Response family link method Marginal Conditional
## 1 simpson gaussian identity none 0.153 0.167Predator-Prey Index
No significant interactions, but best model contains season only.
model.predator.prey.total <-lme(trophic.index~season,
random=list(sample.id=~1, tree.identity=~1), data = nematode)
summary(model.predator.prey.total)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.04046 0.0145 106 2.783 0.00638
## seasonSpring 0.00594 0.0165 106 0.361 0.71885
## seasonSummer 0.02650 0.0170 106 1.559 0.12209
## seasonFall 0.04128 0.0190 38 2.170 0.03634
## seasonWinter.early 0.02611 0.0186 38 1.406 0.16785anova(model.predator.prey.total)## numDF denDF F-value p-value
## (Intercept) 1 104 143.8 <.0001
## season 4 104 1.8 0.138rsquared(model.predator.prey.total) #18## Response family link method Marginal Conditional
## 1 trophic.index gaussian identity none 0.0549 0.178AIC(model.predator.prey.total) #-413## [1] -413AICc(model.predator.prey.total) #-412## [1] -412pairs(emmeans(model.predator.prey.total, "season"))## contrast estimate SE df t.ratio p.value
## Winter.late - Spring -0.0059 0.0165 106 -0.361 0.9960
## Winter.late - Summer -0.0265 0.0170 106 -1.559 0.5270
## Winter.late - Fall -0.0413 0.0190 38 -2.170 0.2130
## Winter.late - Winter.early -0.0261 0.0186 38 -1.406 0.6280
## Spring - Summer -0.0206 0.0130 106 -1.583 0.5120
## Spring - Fall -0.0353 0.0160 38 -2.206 0.1990
## Spring - Winter.early -0.0202 0.0155 38 -1.303 0.6910
## Summer - Fall -0.0148 0.0152 38 -0.970 0.8670
## Summer - Winter.early 0.0004 0.0147 38 0.026 1.0000
## Fall - Winter.early 0.0152 0.0169 38 0.900 0.8950
##
## Degrees-of-freedom method: containment
## P value adjustment: tukey method for comparing a family of 5 estimatesTotal Plant Parasite to Free Living Nematodes
Best explained by season and the interaction between topography and tree functional type.
model.lifestyle.total <-lme(lifestyle.index~season+landscape.position+ mycorrhizal.fungi.type +
season:landscape.position+mycorrhizal.fungi.type:landscape.position,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode)
summary(model.lifestyle.total)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 1.0399 0.543 104 1.916 0.0582
## seasonSpring -0.2316 0.657 104 -0.352 0.7252
## seasonSummer 0.1827 0.619 104 0.295 0.7684
## seasonFall 0.6564 0.686 33 0.957 0.3453
## seasonWinter.early 1.4364 0.662 33 2.170 0.0373
## landscape.positionUphill 1.4238 0.782 33 1.821 0.0777
## mycorrhizal.fungi.typeECM 0.0982 0.362 33 0.271 0.7878
## seasonSpring:landscape.positionUphill 0.5308 0.928 104 0.572 0.5687
## seasonSummer:landscape.positionUphill -0.8423 0.875 104 -0.963 0.3378
## seasonFall:landscape.positionUphill -0.6586 0.974 33 -0.676 0.5036
## seasonWinter.early:landscape.positionUphill -0.4268 0.932 33 -0.458 0.6499
## landscape.positionUphill:mycorrhizal.fungi.typeECM -1.0502 0.502 33 -2.092 0.0442anova(model.lifestyle.total)## numDF denDF F-value p-value
## (Intercept) 1 100 211.7 <.0001
## season 4 100 4.9 0.0012
## landscape.position 1 37 4.9 0.0335
## mycorrhizal.fungi.type 1 37 2.8 0.1004
## season:landscape.position 4 100 1.3 0.2887
## landscape.position:mycorrhizal.fungi.type 1 37 4.4 0.0434rsquared(model.lifestyle.total) #19## Response family link method Marginal Conditional
## 1 lifestyle.index gaussian identity none 0.192 0.192AIC(model.lifestyle.total) #587## [1] 587AICc(model.lifestyle.total) #591## [1] 591Total Channel Ratio (Fungal Feeding : Bacterial Feeding Nematodes)
Best explained by season.
model.channel.total <-lme(channel.index~season,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode)
summary(model.channel.total)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.0397 0.0331 106 1.200 0.23273
## seasonSpring -0.0211 0.0375 106 -0.562 0.57503
## seasonSummer -0.0118 0.0387 106 -0.305 0.76112
## seasonFall 0.1508 0.0432 38 3.487 0.00125
## seasonWinter.early 0.0320 0.0422 38 0.759 0.45246anova(model.channel.total)## numDF denDF F-value p-value
## (Intercept) 1 104 30.61 <.0001
## season 4 104 6.78 1e-04rsquared(model.channel.total) #29## Response family link method Marginal Conditional
## 1 channel.index gaussian identity none 0.189 0.292AIC(model.channel.total) #-166## [1] -166AICc(model.channel.total) #-165## [1] -165
Nematode Indices by Season
Supplemental Figure With other interactions
These are other significant interactions in the models that are highlighted in the results section.
Other Signficant Interactions in Nematode Models
Nematode Community Composition
The matrix used for this was using nematode relative abundance (calculated from Total Nematodes, so accounting for unknown nematodes in the overall community)
#Running ANOVA of NMDS
adonis2(nematode_matrix.rel ~ season * landscape.position * mycorrhizal.fungi.type, data=nematode,
permutations=999, by = "terms")## Permutation test for adonis under reduced model
## Terms added sequentially (first to last)
## Permutation: free
## Number of permutations: 999
##
## adonis2(formula = nematode_matrix.rel ~ season * landscape.position * mycorrhizal.fungi.type, data = nematode, permutations = 999, by = "terms")
## Df SumOfSqs R2 F Pr(>F)
## season 4 0.62 0.145 7.08 0.001 ***
## landscape.position 1 0.22 0.052 10.23 0.001 ***
## mycorrhizal.fungi.type 1 0.02 0.005 0.88 0.356
## season:landscape.position 4 0.27 0.062 3.03 0.018 *
## season:mycorrhizal.fungi.type 4 0.03 0.008 0.39 0.902
## landscape.position:mycorrhizal.fungi.type 1 0.15 0.035 6.87 0.006 **
## season:landscape.position:mycorrhizal.fungi.type 4 0.02 0.004 0.22 0.988
## Residual 135 2.96 0.689
## Total 154 4.29 1.000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Season and landscape position signficant
# landscape position and mycorrhizal fungi
# season and landscape position alone
adonis2(nematode_matrix.rel ~ PRECIP + AIRTEMP + soil.moisture, data=nematode,
permutations=999, by = "terms")## Permutation test for adonis under reduced model
## Terms added sequentially (first to last)
## Permutation: free
## Number of permutations: 999
##
## adonis2(formula = nematode_matrix.rel ~ PRECIP + AIRTEMP + soil.moisture, data = nematode, permutations = 999, by = "terms")
## Df SumOfSqs R2 F Pr(>F)
## PRECIP 1 0.13 0.030 5.01 0.013 *
## AIRTEMP 1 0.18 0.042 6.85 0.004 **
## soil.moisture 1 0.05 0.011 1.75 0.180
## Residual 151 3.94 0.917
## Total 154 4.29 1.000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# precipitation and air temp also drive community composition!
NMDS of Nematode Community Composition
NMDS of Nematode Community Composition by Climate
Macrofauna
Initial models included season, sample type (ecological group), landscape position (topography), and mycorhizal type (tree functional type) + all interactions
Macrofauna Total Abundance
Abundance driven by season, ecological group, and interaction between topography and tree functional type
# model with sample type
bug.total <-lme(sqrt(totalAbundance)~ season + sample.type + season:landscape.position +
mycorrhizal.fungi.type:landscape.position,
random=list(tree.identity=~1, sample.id=~1), data = macrofauna)
summary(bug.total) ## Linear mixed-effects model fit by REML
## Data: macrofauna
## AIC BIC logLik
## 257 278 -113
##
## Random effects:
## Formula: ~1 | tree.identity
## (Intercept)
## StdDev: 0.000192
##
## Formula: ~1 | sample.id %in% tree.identity
## (Intercept) Residual
## StdDev: 0.00227 11
##
## Fixed effects: sqrt(totalAbundance) ~ season + sample.type + season:landscape.position + mycorrhizal.fungi.type:landscape.position
## Value Std.Error DF t-value p-value
## (Intercept) 12.85 6.27 16 2.05 0.0572
## seasonSpring 11.15 7.78 2 1.43 0.2882
## seasonSummer 16.48 7.78 2 2.12 0.1682
## seasonFall 5.97 7.78 16 0.77 0.4536
## seasonWinter.early 4.14 7.78 16 0.53 0.6017
## sample.typeSoil 23.47 3.48 16 6.75 0.0000
## seasonWinter.late:landscape.positionUphill -4.98 8.61 2 -0.58 0.6211
## seasonSpring:landscape.positionUphill -16.07 8.61 2 -1.87 0.2030
## seasonSummer:landscape.positionUphill -6.61 8.16 9 -0.81 0.4384
## seasonFall:landscape.positionUphill 11.73 8.61 16 1.36 0.1917
## seasonWinter.early:landscape.positionUphill -11.55 8.61 16 -1.34 0.1982
## landscape.positionDownhill:mycorrhizal.fungi.typeECM -7.84 4.92 2 -1.59 0.2520
## landscape.positionUphill:mycorrhizal.fungi.typeECM 13.31 5.50 2 2.42 0.1366
## Correlation:
## (Intr) ssnSpr ssnSmm ssnFll ssnWn. smpl.S ssnWntr.l:.U ssnSp:.U ssnSm:.U ssF:.U ssnWntr.r:.U l.D:..
## seasonSpring -0.620
## seasonSummer -0.620 0.500
## seasonFall -0.620 0.500 0.500
## seasonWinter.early -0.620 0.500 0.500 0.500
## sample.typeSoil -0.277 0.000 0.000 0.000 0.000
## seasonWinter.late:landscape.positionUphill -0.672 0.452 0.452 0.452 0.452 0.000
## seasonSpring:landscape.positionUphill -0.112 -0.452 0.000 0.000 0.000 0.000 0.184
## seasonSummer:landscape.positionUphill -0.118 0.000 -0.477 0.000 0.000 0.000 0.086 0.086
## seasonFall:landscape.positionUphill -0.112 0.000 0.000 -0.452 0.000 0.000 0.184 0.184 0.086
## seasonWinter.early:landscape.positionUphill -0.112 0.000 0.000 0.000 -0.452 0.000 0.184 0.184 0.086 0.184
## landscape.positionDownhill:mycorrhizal.fungi.typeECM -0.392 0.000 0.000 0.000 0.000 0.000 0.286 0.286 0.302 0.286 0.286
## landscape.positionUphill:mycorrhizal.fungi.typeECM 0.000 0.000 0.000 0.000 0.000 0.000 -0.319 -0.319 0.000 -0.319 -0.319 0.000
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.0483 -0.6084 0.0194 0.3546 2.1949
##
## Number of Observations: 40
## Number of Groups:
## tree.identity sample.id %in% tree.identity
## 11 19anova(bug.total)## numDF denDF F-value p-value
## (Intercept) 1 11 299.3 <.0001
## season 4 11 2.8 0.0784
## sample.type 1 11 45.5 <.0001
## season:landscape.position 5 11 1.9 0.1786
## landscape.position:mycorrhizal.fungi.type 2 6 4.2 0.0724rsquared(bug.total)## Response family link method Marginal Conditional
## 1 totalAbundance gaussian identity none 0.657 0.657AIC(bug.total) #257## [1] 257AICc(bug.total) #281## [1] 281pairs(emmeans(bug.total, specs= ~season))## NOTE: A nesting structure was detected in the fitted model:
## landscape.position %in% season, mycorrhizal.fungi.type %in% (season*landscape.position)## NOTE: Results may be misleading due to involvement in interactions## contrast estimate SE df t.ratio p.value
## Winter.late - Spring -5.60 5.5 2 -1.019 0.8340
## Winter.late - Summer -12.34 5.5 2 -2.244 0.4110
## Winter.late - Fall -14.33 5.5 2 -2.606 0.3340
## Winter.late - Winter.early -0.86 5.5 2 -0.156 1.0000
## Spring - Summer -6.73 5.5 2 -1.224 0.7520
## Spring - Fall -8.73 5.5 2 -1.587 0.6110
## Spring - Winter.early 4.75 5.5 2 0.863 0.8900
## Summer - Fall -1.99 5.5 2 -0.362 0.9930
## Summer - Winter.early 11.48 5.5 2 2.088 0.4510
## Fall - Winter.early 13.48 5.5 16 2.450 0.1520
##
## Results are averaged over the levels of: sample.type, mycorrhizal.fungi.type, landscape.position
## Note: contrasts are still on the sqrt scale. Consider using
## regrid() if you want contrasts of back-transformed estimates.
## Degrees-of-freedom method: containment
## P value adjustment: tukey method for comparing a family of 5 estimates
Total Macrofauna Abundance By Season
Macrofauna Shannon Diversity
Driven entirely by ecological group
bug.shan <-lme(shannon~ sample.type,
random=list(tree.identity=~1, sample.id=~1), data = macrofauna)
summary(bug.shan)## Linear mixed-effects model fit by REML
## Data: macrofauna
## AIC BIC logLik
## 41.9 50.1 -16
##
## Random effects:
## Formula: ~1 | tree.identity
## (Intercept)
## StdDev: 0.189
##
## Formula: ~1 | sample.id %in% tree.identity
## (Intercept) Residual
## StdDev: 0.0948 0.294
##
## Fixed effects: shannon ~ sample.type
## Value Std.Error DF t-value p-value
## (Intercept) 1.311 0.0916 20 14.32 0.0000
## sample.typeSoil 0.325 0.0930 20 3.49 0.0023
## Correlation:
## (Intr)
## sample.typeSoil -0.508
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.9796 -0.4680 0.0439 0.5497 2.2599
##
## Number of Observations: 40
## Number of Groups:
## tree.identity sample.id %in% tree.identity
## 11 19anova(bug.shan)## numDF denDF F-value p-value
## (Intercept) 1 20 349 <.0001
## sample.type 1 20 12 0.0023rsquared(bug.shan)## Response family link method Marginal Conditional
## 1 shannon gaussian identity none 0.171 0.453AIC(bug.shan) #42 ## [1] 41.9AICc(bug.shan) #44## [1] 43.7
Composition by Sample Type
Macrofauna By Sample Type
Models included season, topography, and tree functional type + all interactions
Epigeic Abundance
No significant interactions in any model - this is the best using AIC
litter.total <-lme(sqrt(totalAbundance)~ landscape.position + mycorrhizal.fungi.type +
mycorrhizal.fungi.type:landscape.position,
random=list(tree.identity=~1, sample.id=~1), data = macrofauna[macrofauna$sample.type == "Litter",])
summary(litter.total)## Linear mixed-effects model fit by REML
## Data: macrofauna[macrofauna$sample.type == "Litter", ]
## AIC BIC logLik
## 144 149 -64.9
##
## Random effects:
## Formula: ~1 | tree.identity
## (Intercept)
## StdDev: 8.59
##
## Formula: ~1 | sample.id %in% tree.identity
## (Intercept) Residual
## StdDev: 0.00108 9.34
##
## Fixed effects: sqrt(totalAbundance) ~ landscape.position + mycorrhizal.fungi.type + mycorrhizal.fungi.type:landscape.position
## Value Std.Error DF t-value p-value
## (Intercept) 17.44 6.54 9 2.666 0.0258
## landscape.positionUphill 5.18 8.10 6 0.639 0.5462
## mycorrhizal.fungi.typeECM -3.28 8.86 9 -0.370 0.7200
## landscape.positionUphill:mycorrhizal.fungi.typeECM -0.44 11.42 6 -0.038 0.9707
## Correlation:
## (Intr) lnds.U m..ECM
## landscape.positionUphill -0.714
## mycorrhizal.fungi.typeECM -0.739 0.527
## landscape.positionUphill:mycorrhizal.fungi.typeECM 0.507 -0.709 -0.614
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -1.219 -0.427 -0.146 0.187 2.081
##
## Number of Observations: 20
## Number of Groups:
## tree.identity sample.id %in% tree.identity
## 11 19anova(litter.total)## numDF denDF F-value p-value
## (Intercept) 1 9 29.02 0.0004
## landscape.position 1 6 0.93 0.3721
## mycorrhizal.fungi.type 1 9 0.25 0.6303
## landscape.position:mycorrhizal.fungi.type 1 6 0.00 0.9707rsquared(litter.total)## Response family link method Marginal Conditional
## 1 totalAbundance gaussian identity none 0.0619 0.492AIC(litter.total) #144## [1] 144AICc(litter.total) #153## [1] 153Epigeic Diversity (Shannon)
Topography weakly trending towards significant (p=0.08)
litter.shan <-lme(shannon~ landscape.position,
random=list(tree.identity=~1, sample.id=~1), data = macrofauna[macrofauna$sample.type == "Litter",])
summary(litter.shan)## Linear mixed-effects model fit by REML
## Data: macrofauna[macrofauna$sample.type == "Litter", ]
## AIC BIC logLik
## 29 33.4 -9.48
##
## Random effects:
## Formula: ~1 | tree.identity
## (Intercept)
## StdDev: 0.464
##
## Formula: ~1 | sample.id %in% tree.identity
## (Intercept) Residual
## StdDev: 0.203 0.075
##
## Fixed effects: shannon ~ landscape.position
## Value Std.Error DF t-value p-value
## (Intercept) 1.464 0.170 10 8.61 0.0000
## landscape.positionUphill -0.343 0.171 7 -2.01 0.0849
## Correlation:
## (Intr)
## landscape.positionUphill -0.475
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -0.6168 -0.1412 -0.0257 0.2080 0.7797
##
## Number of Observations: 20
## Number of Groups:
## tree.identity sample.id %in% tree.identity
## 11 19anova(litter.shan)## numDF denDF F-value p-value
## (Intercept) 1 10 75.6 <.0001
## landscape.position 1 7 4.0 0.0849rsquared(litter.shan)## Response family link method Marginal Conditional
## 1 shannon gaussian identity none 0.106 0.981AIC(litter.shan) #29## [1] 29AICc(litter.shan) #33## [1] 33.2Endogeic Abundance
Interaction (topography x tree functional type) is trending towards significant (p=0.08)
soil.total <-lme(sqrt(totalAbundance)~ mycorrhizal.fungi.type + mycorrhizal.fungi.type:landscape.position,
random=list(tree.identity=~1, sample.id=~1), data = macrofauna[macrofauna$sample.type == "Soil",])
summary(soil.total)## Linear mixed-effects model fit by REML
## Data: macrofauna[macrofauna$sample.type == "Soil", ]
## AIC BIC logLik
## 149 154 -67.5
##
## Random effects:
## Formula: ~1 | tree.identity
## (Intercept)
## StdDev: 4.88e-05
##
## Formula: ~1 | sample.id %in% tree.identity
## (Intercept) Residual
## StdDev: 0.00321 13.4
##
## Fixed effects: sqrt(totalAbundance) ~ mycorrhizal.fungi.type + mycorrhizal.fungi.type:landscape.position
## Value Std.Error DF t-value p-value
## (Intercept) 47.5 6.01 9 7.90 0.0000
## mycorrhizal.fungi.typeECM -13.3 8.50 9 -1.57 0.1518
## mycorrhizal.fungi.typeAM:landscape.positionUphill -12.9 8.14 6 -1.58 0.1651
## mycorrhizal.fungi.typeECM:landscape.positionUphill 20.9 9.02 6 2.32 0.0594
## Correlation:
## (Intr) my..ECM m..AM:
## mycorrhizal.fungi.typeECM -0.707
## mycorrhizal.fungi.typeAM:landscape.positionUphill -0.739 0.522
## mycorrhizal.fungi.typeECM:landscape.positionUphill 0.000 -0.471 0.000
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -1.407 -0.667 -0.255 0.527 2.190
##
## Number of Observations: 20
## Number of Groups:
## tree.identity sample.id %in% tree.identity
## 11 19anova(soil.total)## numDF denDF F-value p-value
## (Intercept) 1 9 193.5 <.0001
## mycorrhizal.fungi.type 1 9 0.2 0.6309
## mycorrhizal.fungi.type:landscape.position 2 6 3.9 0.0807rsquared(soil.total)## Response family link method Marginal Conditional
## 1 totalAbundance gaussian identity none 0.3 0.3AIC(soil.total) #149## [1] 149AICc(soil.total) #158## [1] 158pairs(emmeans(soil.total, specs= ~landscape.position*mycorrhizal.fungi.type))## NOTE: A nesting structure was detected in the fitted model:
## landscape.position %in% mycorrhizal.fungi.type## contrast estimate SE df t.ratio p.value
## Downhill AM - Uphill AM 12.87 8.14 6 1.580 0.4530
## Downhill AM - Downhill ECM 13.32 8.50 9 1.566 0.4420
## Downhill AM - Uphill ECM -7.62 9.02 6 -0.844 0.8320
## Uphill AM - Downhill ECM 0.45 8.14 6 0.055 1.0000
## Uphill AM - Uphill ECM -20.49 8.68 6 -2.360 0.1860
## Downhill ECM - Uphill ECM -20.93 9.02 6 -2.321 0.1950
##
## Note: contrasts are still on the sqrt scale. Consider using
## regrid() if you want contrasts of back-transformed estimates.
## Degrees-of-freedom method: containment
## P value adjustment: tukey method for comparing a family of 4 estimatesEndogeic Shannon Diversity
Not signficant, best model is null model
soil.shan <-lme(shannon~ landscape.position,
random=list(tree.identity=~1, sample.id=~1), data = macrofauna[macrofauna$sample.type == "Soil",])
summary(soil.shan)## Linear mixed-effects model fit by REML
## Data: macrofauna[macrofauna$sample.type == "Soil", ]
## AIC BIC logLik
## 7.65 12.1 1.18
##
## Random effects:
## Formula: ~1 | tree.identity
## (Intercept)
## StdDev: 7.97e-07
##
## Formula: ~1 | sample.id %in% tree.identity
## (Intercept) Residual
## StdDev: 0.241 0.006
##
## Fixed effects: shannon ~ landscape.position
## Value Std.Error DF t-value p-value
## (Intercept) 1.571 0.0762 10 20.63 0.000
## landscape.positionUphill 0.156 0.1107 7 1.41 0.201
## Correlation:
## (Intr)
## landscape.positionUphill -0.688
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -0.70454 -0.01408 0.00384 0.01607 0.70949
##
## Number of Observations: 20
## Number of Groups:
## tree.identity sample.id %in% tree.identity
## 11 19anova(soil.shan)## numDF denDF F-value p-value
## (Intercept) 1 10 886 <.0001
## landscape.position 1 7 2 0.201rsquared(soil.shan)## Response family link method Marginal Conditional
## 1 shannon gaussian identity none 0.0997 0.999AIC(soil.shan) #11## [1] 7.65AICc(soil.shan) #18## [1] 11.9Macrofauna Community Composition
# Running ANOVA of NMDS
adonis2(macrofauna_matrix.rel ~ season * landscape.position * mycorrhizal.fungi.type * sample.type,
data=macrofauna, permutations=999, by = "terms")## Permutation test for adonis under reduced model
## Terms added sequentially (first to last)
## Permutation: free
## Number of permutations: 999
##
## adonis2(formula = macrofauna_matrix.rel ~ season * landscape.position * mycorrhizal.fungi.type * sample.type, data = macrofauna, permutations = 999, by = "terms")
## Df SumOfSqs R2 F Pr(>F)
## season 4 1.73 0.189 2.84 0.033 *
## landscape.position 1 0.36 0.040 2.38 0.102
## mycorrhizal.fungi.type 1 0.12 0.013 0.80 0.605
## sample.type 1 2.04 0.223 13.41 0.001 ***
## season:landscape.position 4 0.49 0.053 0.80 0.687
## season:mycorrhizal.fungi.type 4 0.86 0.094 1.41 0.250
## landscape.position:mycorrhizal.fungi.type 1 0.21 0.023 1.41 0.274
## season:sample.type 4 0.99 0.108 1.63 0.194
## landscape.position:sample.type 1 0.24 0.027 1.60 0.229
## mycorrhizal.fungi.type:sample.type 1 0.10 0.011 0.64 0.689
## season:landscape.position:mycorrhizal.fungi.type 3 0.48 0.052 1.05 0.475
## season:landscape.position:sample.type 4 0.52 0.057 0.85 0.632
## season:mycorrhizal.fungi.type:sample.type 4 0.32 0.035 0.52 0.928
## landscape.position:mycorrhizal.fungi.type:sample.type 1 0.13 0.014 0.86 0.538
## season:landscape.position:mycorrhizal.fungi.type:sample.type 3 0.24 0.026 0.53 0.898
## Residual 2 0.30 0.033
## Total 39 9.14 1.000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# only season and sample type are significant :) (no interaction)
# what about with season combined?
adonis2(macrofauna_matrix.rel ~ season.1 * landscape.position * mycorrhizal.fungi.type * sample.type,
data=macrofauna, permutations=999, by = "terms")## Permutation test for adonis under reduced model
## Terms added sequentially (first to last)
## Permutation: free
## Number of permutations: 999
##
## adonis2(formula = macrofauna_matrix.rel ~ season.1 * landscape.position * mycorrhizal.fungi.type * sample.type, data = macrofauna, permutations = 999, by = "terms")
## Df SumOfSqs R2 F Pr(>F)
## season.1 3 1.36 0.149 2.56 0.008 **
## landscape.position 1 0.36 0.040 2.05 0.065 .
## mycorrhizal.fungi.type 1 0.12 0.013 0.69 0.674
## sample.type 1 2.04 0.223 11.50 0.001 ***
## season.1:landscape.position 3 0.40 0.044 0.75 0.741
## season.1:mycorrhizal.fungi.type 3 0.65 0.071 1.21 0.269
## landscape.position:mycorrhizal.fungi.type 1 0.21 0.023 1.21 0.290
## season.1:sample.type 3 0.86 0.094 1.61 0.081 .
## landscape.position:sample.type 1 0.24 0.027 1.37 0.225
## mycorrhizal.fungi.type:sample.type 1 0.10 0.011 0.55 0.763
## season.1:landscape.position:mycorrhizal.fungi.type 2 0.25 0.027 0.70 0.736
## season.1:landscape.position:sample.type 3 0.38 0.042 0.72 0.762
## season.1:mycorrhizal.fungi.type:sample.type 3 0.11 0.012 0.21 0.999
## landscape.position:mycorrhizal.fungi.type:sample.type 1 0.13 0.014 0.74 0.626
## season.1:landscape.position:mycorrhizal.fungi.type:sample.type 2 0.15 0.016 0.41 0.947
## Residual 10 1.77 0.194
## Total 39 9.14 1.000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# pretty much the same - weak relationship w fungi (p=0.1) and interaction between season and sample type (p=0.09)
adonis2(macrofauna_matrix.rel ~ PRECIP + AIRTEMP + soil.moisture,
data=macrofauna, permutations=999, by = "terms", na.action = na.omit)## Permutation test for adonis under reduced model
## Terms added sequentially (first to last)
## Permutation: free
## Number of permutations: 999
##
## adonis2(formula = macrofauna_matrix.rel ~ PRECIP + AIRTEMP + soil.moisture, data = macrofauna, permutations = 999, by = "terms", na.action = na.omit)
## Df SumOfSqs R2 F Pr(>F)
## PRECIP 1 0.80 0.106 4.08 0.001 ***
## AIRTEMP 1 0.28 0.037 1.43 0.217
## soil.moisture 1 0.60 0.079 3.03 0.014 *
## Residual 30 5.89 0.778
## Total 33 7.57 1.000
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Precip and soil moisture significant
NMDS of Macrofauna Community Composition
NMDS of Macrofauna with Climate Variables
Within Season Nematode Analyses
What are the variables driving changes in nematode community composition within each season?
Used as explanatory, but not highlighted in the results section
Predator-Prey Index
Winter Predator-Prey Index
model.predator.prey.winter <-lme(trophic.index~ PRECIP,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.winter)
summary(model.predator.prey.winter)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.075142 0.007123 29 10.55 1.94e-11
## PRECIP -0.000528 0.000143 29 -3.69 9.15e-04anova(model.predator.prey.winter)## numDF denDF F-value p-value
## (Intercept) 1 29 126.5 <.0001
## PRECIP 1 29 13.6 9e-04rsquared(model.predator.prey.winter) #32## Response family link method Marginal Conditional
## 1 trophic.index gaussian identity none 0.245 0.323AIC(model.predator.prey.winter) #-167## [1] -163AICc(model.predator.prey.winter) #-166## [1] -161Spring Predator-Prey Index
model.predator.prey.spring <-lme(trophic.index~ PRECIP,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.spring)
summary(model.predator.prey.spring)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.056309 0.007280 16 7.74 8.56e-07
## PRECIP -0.000421 0.000195 10 -2.15 5.67e-02anova(model.predator.prey.spring)## numDF denDF F-value p-value
## (Intercept) 1 16 69.0 <.0001
## PRECIP 1 10 4.6 0.0567rsquared(model.predator.prey.spring) #92## Response family link method Marginal Conditional
## 1 trophic.index gaussian identity none 0.0651 0.682AIC(model.predator.prey.spring) #-121## [1] -115AICc(model.predator.prey.spring) #-118## [1] -113Summer Predator-Prey Index
model.predator.prey.summer <-lme(trophic.index~ AIRTEMP + landscape.position,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.summer)
summary(model.predator.prey.summer)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.08409 0.06799 31 1.237 0.225
## AIRTEMP -0.00176 0.00303 31 -0.581 0.566
## landscape.positionUphill 0.04292 0.01488 14 2.884 0.012anova(model.predator.prey.summer)## numDF denDF F-value p-value
## (Intercept) 1 31 80.0 <.0001
## AIRTEMP 1 31 0.3 0.566
## landscape.position 1 14 8.3 0.012rsquared(model.predator.prey.summer) #36## Response family link method Marginal Conditional
## 1 trophic.index gaussian identity none 0.209 0.374AIC(model.predator.prey.summer) #-140## [1] -135AICc(model.predator.prey.summer) #-137## [1] -133Fall Predator-Prey Index
Best model was null model, but this is the next best - precip is not significant
model.predator.prey.fall <-lme(trophic.index~ PRECIP,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.fall)
summary(model.predator.prey.fall)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.07324 0.02672 18 2.741 0.0134
## PRECIP 0.00145 0.00241 18 0.603 0.5538anova(model.predator.prey.fall)## numDF denDF F-value p-value
## (Intercept) 1 18 14.51 0.0013
## PRECIP 1 18 0.36 0.5538rsquared(model.predator.prey.fall) #11## Response family link method Marginal Conditional
## 1 trophic.index gaussian identity none 0.0131 0.0685AIC(model.predator.prey.fall) #-22## [1] -20.8AICc(model.predator.prey.fall) #-19## [1] -18Nematode Lifestyle Index
This index is a ratio of plant parasite vs free living nematodes
Winter Lifestyle Index
No significant variables
model.lifestyle.winter <-lme(lifestyle.index~ soil.moisture + mycorrhizal.fungi.type + landscape.position:mycorrhizal.fungi.type,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.winter)
summary(model.lifestyle.winter)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 6.488 3.065 29 2.117 0.043
## soil.moisture -0.152 0.101 29 -1.504 0.143
## mycorrhizal.fungi.typeECM -0.176 0.950 12 -0.185 0.856
## mycorrhizal.fungi.typeAM:landscape.positionUphill 1.211 0.890 12 1.361 0.199
## mycorrhizal.fungi.typeECM:landscape.positionUphill -0.345 0.972 12 -0.355 0.729anova(model.lifestyle.winter)## numDF denDF F-value p-value
## (Intercept) 1 29 51.7 <.0001
## soil.moisture 1 29 1.7 0.199
## mycorrhizal.fungi.type 1 12 1.9 0.189
## mycorrhizal.fungi.type:landscape.position 2 12 1.0 0.397rsquared(model.lifestyle.winter) #20## Response family link method Marginal Conditional
## 1 lifestyle.index gaussian identity none 0.113 0.148AIC(model.lifestyle.winter) #219## [1] 219AICc(model.lifestyle.winter) #223## [1] 223Spring Lifestyle Index
model.lifestyle.spring <-lme(lifestyle.index~ landscape.position,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.spring)
summary(model.lifestyle.spring)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.845 0.308 15 2.75 0.01493
## landscape.positionUphill 1.442 0.435 15 3.32 0.00471anova(model.lifestyle.spring)## numDF denDF F-value p-value
## (Intercept) 1 15 51.9 <.0001
## landscape.position 1 15 11.0 0.0047rsquared(model.lifestyle.spring) #26## Response family link method Marginal Conditional
## 1 lifestyle.index gaussian identity none 0.262 0.262AIC(model.lifestyle.spring) #113## [1] 113AICc(model.lifestyle.spring) #115## [1] 115Summer Lifestyle Index
model.lifestyle.summer <-lme(lifestyle.index~ PRECIP,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.summer)
summary(model.lifestyle.summer)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 1.57822 0.18939 31 8.33 2.06e-09
## PRECIP -0.00582 0.00303 31 -1.92 6.35e-02anova(model.lifestyle.summer)## numDF denDF F-value p-value
## (Intercept) 1 31 83.2 <.0001
## PRECIP 1 31 3.7 0.0635rsquared(model.lifestyle.summer) #36## Response family link method Marginal Conditional
## 1 lifestyle.index gaussian identity none 0.0496 0.371AIC(model.lifestyle.summer) #127## [1] 127AICc(model.lifestyle.summer) #128## [1] 128Fall Lifestyle Index
Best model is null - soil moisture not signficant and rsquared is hideous (best r-squared value from stepwise is from the full model)
model.lifestyle.fall <-lme(lifestyle.index~ soil.moisture,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.fall)
summary(model.lifestyle.fall)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 1.95820 0.8032 18 2.438 0.0254
## soil.moisture -0.00537 0.0308 18 -0.174 0.8635anova(model.lifestyle.fall)## numDF denDF F-value p-value
## (Intercept) 1 18 33.3 <.0001
## soil.moisture 1 18 0.0 0.864rsquared(model.lifestyle.fall) #.5## Response family link method Marginal Conditional
## 1 lifestyle.index gaussian identity none 0.00117 0.00117AIC(model.lifestyle.fall) #117## [1] 117AICc(model.lifestyle.fall) #120## [1] 120Nematode Channel Index
Ratio of fungal feeding and bacterial feeding nematodes ### Winter Channel Index
model.channel.winter <-lme(channel.index~ soil.moisture,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.winter)
summary(model.channel.winter)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.27779 0.09491 29 2.93 0.0066
## soil.moisture -0.00768 0.00334 29 -2.30 0.0289anova(model.channel.winter)## numDF denDF F-value p-value
## (Intercept) 1 29 30.07 <.0001
## soil.moisture 1 29 5.29 0.0289rsquared(model.channel.winter) #10## Response family link method Marginal Conditional
## 1 channel.index gaussian identity none 0.101 0.101AIC(model.channel.winter) #-85## [1] -85AICc(model.channel.winter) #-84## [1] -83.6Spring Channel Index
model.channel.spring <-lme(channel.index~ landscape.position,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.spring)
summary(model.channel.spring)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.0391 0.0127 15 3.08 0.00765
## landscape.positionUphill -0.0356 0.0176 15 -2.02 0.06127anova(model.channel.spring)## numDF denDF F-value p-value
## (Intercept) 1 15 5.46 0.0337
## landscape.position 1 15 4.09 0.0613rsquared(model.channel.spring) #70## Response family link method Marginal Conditional
## 1 channel.index gaussian identity none 0.173 0.7AIC(model.channel.spring) #-102## [1] -102AICc(model.channel.spring) #-100## [1] -99.4Summer Channel Index
Best model is null model - no significant predictor variables
model.channel.summer <-lme(channel.index~ AIRTEMP,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.summer)
summary(model.channel.summer)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.05858 0.07745 31 0.756 0.455
## AIRTEMP -0.00124 0.00344 31 -0.361 0.720anova(model.channel.summer)## numDF denDF F-value p-value
## (Intercept) 1 31 4.51 0.0418
## AIRTEMP 1 31 0.13 0.7204rsquared(model.channel.summer) #61## Response family link method Marginal Conditional
## 1 channel.index gaussian identity none 0.00113 0.601AIC(model.channel.summer) #-116## [1] -115AICc(model.channel.summer) #-115## [1] -114Fall Channel Index
Best model is null, no signficant predictor variables
model.channel.fall <-lme(channel.index~AIRTEMP,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.fall)
summary(model.channel.fall)$tTable## Value Std.Error DF t-value p-value
## (Intercept) -0.2159 0.3182 18 -0.679 0.506
## AIRTEMP 0.0244 0.0188 18 1.299 0.210anova(model.channel.fall)## numDF denDF F-value p-value
## (Intercept) 1 18 11.19 0.0036
## AIRTEMP 1 18 1.69 0.2105rsquared(model.channel.fall) #17## Response family link method Marginal Conditional
## 1 channel.index gaussian identity none 0.0558 0.154AIC(model.channel.fall) #22## [1] 22.4AICc(model.channel.fall) #25## [1] 25.2Nematode Diversity (Simpson Diversity Index)
Winter Nematode Diversity
model.diversity.winter <-lme(simpson~ soil.moisture,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.winter)
summary(model.diversity.winter)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.24122 0.13369 29 1.80 0.0816
## soil.moisture 0.00811 0.00469 29 1.73 0.0946anova(model.diversity.winter) ## numDF denDF F-value p-value
## (Intercept) 1 29 791 <.0001
## soil.moisture 1 29 3 0.0946rsquared(model.diversity.winter) #25## Response family link method Marginal Conditional
## 1 simpson gaussian identity none 0.0646 0.166AIC(model.diversity.winter) #-60## [1] -58.4AICc(model.diversity.winter) #-59## [1] -57Spring Nematode Diversity
model.diversity.spring <-lme(simpson~ soil.moisture + landscape.position,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.spring)
summary(model.diversity.spring)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.72258 0.1079 15 6.70 7.11e-06
## soil.moisture -0.00735 0.0039 10 -1.89 8.86e-02
## landscape.positionUphill -0.06591 0.0268 15 -2.46 2.66e-02anova(model.diversity.spring)## numDF denDF F-value p-value
## (Intercept) 1 15 1389 <.0001
## soil.moisture 1 10 2 0.1603
## landscape.position 1 15 6 0.0266rsquared(model.diversity.spring) #28## Response family link method Marginal Conditional
## 1 simpson gaussian identity none 0.212 0.212AIC(model.diversity.spring) #-46## [1] -44.6AICc(model.diversity.spring) #-43## [1] -41.2Summer Nematode Diversity
Best model is null, no significant variables - but in worse AIC model, AIRTEMP trends towards signficant (May rerun this or ask input!)
model.diversity.summer <-lme(simpson~ mycorrhizal.fungi.type,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.summer)
summary(model.diversity.summer)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.5146 0.0177 32 29.083 1.39e-24
## mycorrhizal.fungi.typeECM 0.0166 0.0271 14 0.614 5.49e-01anova(model.diversity.summer)## numDF denDF F-value p-value
## (Intercept) 1 32 1518 <.0001
## mycorrhizal.fungi.type 1 14 0 0.549rsquared(model.diversity.summer) #30## Response family link method Marginal Conditional
## 1 simpson gaussian identity none 0.0127 0.295AIC(model.diversity.summer) #-99## [1] -98.7AICc(model.diversity.summer) #-97## [1] -97.3Fall Nematode Diversity
Best model is null, no signficant predictor variables
model.diversity.fall <-lme(simpson~ soil.moisture,
random=list(sample.id=~1, tree.identity=~1), na.action = na.omit, data = nematode.fall)
summary(model.diversity.fall)$tTable## Value Std.Error DF t-value p-value
## (Intercept) 0.511375 0.05826 18 8.778 6.38e-08
## soil.moisture 0.000523 0.00224 18 0.234 8.18e-01anova(model.diversity.fall)## numDF denDF F-value p-value
## (Intercept) 1 18 518 <.0001
## soil.moisture 1 18 0 0.818rsquared(model.diversity.fall) #1## Response family link method Marginal Conditional
## 1 simpson gaussian identity none 0.0021 0.0021AIC(model.diversity.fall) #-14## [1] -14AICc(model.diversity.fall) #-11## [1] -11.1